Deep Learning and Common Sensitivity

OpenAI's language model GPT-3 makes things up that don't make any sense. So why is it so hard for a computer, which is being programmed by humans, to mimic? Yeh gen and her colleagues built one of the first common sense knowledge graphs. They combined it with a neural network that could learn from that textbook. The result was called comet. When researchers typed a real world common sense problem into comet, its guesses were 70% plausible. And resembles how humans learn.